Depth Prediction Without the Sensors: Leveraging Structure for Unsupervised Learning from Monocular Videos | Read Paper on Bytez